They’re very sneaky about their science. I remember reading one article that claimed that 97% of transgenders were happy with their results. Turns out that was based on a mail-in survey, and almost half of the recipients didn’t respond and may have even been dead. They meant to say, “97% of those who responded, which was only about 60%,” but I guess they forgot, eh?
Statistics become weapons in the hands of pathological liars.
No statistical analysis of anything can be trusted without an in depth understanding of both the methodology and the subject being discussed.
When in doubt my default position is that everybody who uses statistics is lying.